A new, timely and detailed report outlines the rise of ‘embodied AI’ and maps the companies best positioned to lead the humanoid robotics revolution.
Humanoid robots could drive a $3 trillion total addressable market (TAM) in the United States alone, according to a new report by Morgan Stanley.
The global financial services company describes the emerging field as “embodied AI” – a combination of generative artificial intelligence and robotics hardware that may soon transform how labor is delivered across industries.
The deep-dive report, titled The Humanoid 100, details the companies that Morgan Stanley believes are best positioned to benefit from the shift toward physical AI.
With interest in humanoids accelerating, particularly following Nvidia CEO Jensen Huang’s high-profile CES 2025 presentation on robotics, the bank argues that the theme is becoming too large to ignore.
“The physical embodiment of AI touches a $60 trillion TAM, global GDP, and the meaning of work,” the report states.
While previous discussions around AI were largely software-centric, the bank notes that generative AI is beginning to migrate from the digital world of bits and bytes to the physical world of atoms and photons.
Humanoids – robots with humanlike form and motion – are now viewed as a serious frontier in automation, with potential applications ranging from manufacturing and logistics to home services, elder care, and defense.
A roadmap for investors
Morgan Stanley has compiled a list of 100 public companies – dubbed the “Humanoid 100” – with material exposure to the humanoid value chain. The list is broken into three main segments:
- Brain: AI models, semiconductors, vision systems, and digital twin simulation software
- Body: Sensors, actuators, rare-earth magnets, motors, bearings, and power components
- Integrators: Companies building or expected to build complete humanoid robots
Of the 100 companies, 52 percent are reported to be actively involved in humanoid development. The remainder are either close competitors of involved firms or are believed to have the technical and commercial potential to enter the field.
Among the notable names included are Nvidia (AI models and chips), Tesla (developer of the Optimus robot), Siemens (automation systems), Rockwell Automation, ABB, Apple, and Amazon. The report also highlights component specialists such as Nidec (motors), MP Materials (rare-earths), and Hexagon (vision and simulation).
Asia leads, the West lags
One of the report’s most striking findings is the geographic imbalance in the humanoid sector. A majority – 73 percent – of the companies involved are based in Asia, and 56 percent are in China alone. Only a handful of US or European companies appear to be building full humanoid robots at this stage.
“A common refrain we hear from investors is the lack of Western firms to add to their humanoid portfolio outside of TSLA and NVDA,” the report notes.
The researchers argue that this imbalance could mirror the experience of the electric vehicle sector, where China has taken a dominant position. In both cases, overlapping supply chains and strong national support have enabled Chinese startups to scale quickly.
Morgan Stanley suggests that policymakers and investors in the West may need to respond strategically if they hope to catch up.
Dissecting the humanoid
The report devotes significant space to the technical architecture of a humanoid robot. The “brain” consists of foundational AI models like Nvidia’s Project Gr00t, high-performance compute chips, and software for simulation and control. The “body” comprises an array of parts including:
- Sensors: Vision, lidar, force, torque, and magnetic field detectors
- Actuators: Motors and mechanisms that enable movement, often with 16-60 degrees of freedom
- Power and structure: Lithium-ion batteries, thermal management systems, and lightweight alloys
Many of the parts used in humanoids are shared with the EV or industrial automation industries, giving companies in those sectors a potential head start. However, building a functioning humanoid remains complex and expensive, with current designs typically optimized for demonstration and research.
Future of work, or future of investment?
While the path to mass adoption is likely to take decades, Morgan Stanley expects commercial deployment to begin in earnest in the 2030s, reaching global scale by 2050. The US market alone could support $3 trillion in annual economic activity related to humanoid robots, the report estimates.
This would represent a radical redefinition of labor and productivity, especially if humanoids prove viable in areas currently limited by human workforce availability. From warehouse work and elder care to hazardous industrial tasks, embodied AI could reshape not only businesses but social structures.
Yet the report stops short of declaring winners, and acknowledges that many developments will remain unpredictable.
“We do not pretend to present the ‘Humanoid 100’ as an exhaustive list,” the authors write. “Rather, (it is) a starting point in a conversation that will last many years and is sure to follow some exciting and unpredictable chapters along the way.”
Defining technology
For now, the humanoid sector remains early and uneven. But with a growing chorus of investors, technologists, and executives taking the idea seriously, Morgan Stanley’s $3 trillion projection could quickly become a central figure in one of the defining technological narratives of the 21st century.
As physical AI marches toward reality, the question may no longer be if humanoids will change the world – but who will build them, and how fast.
